bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

1 The Rpi-mcq1 resistance gene family recognizes Avr2 of

2 Phytophthora infestans but is distinct from R2 3 4 5 Carolina Aguilera-Galvez1, Zhaohui Chu2,a, Sumaiya Haque Omy1,b, Doret Wouters1, 6 Eleanor M. Gilroy3, Jack H. Vossen1, Richard G.F Visser1, Paul Birch3, Jonathan D.G 7 Jones2,& and Vivianne G.A.A Vleeshouwers1,&,* 8 9 10 1Plant Breeding, Wageningen University and Research, Droevendaalsesteeg 1, 6708 PB, 11 Wageningen, The Netherlands. 12 2The Sainsbury Laboratory, University of East Anglia, Norwich Research Park, NR4 13 7UH Norwich, United Kingdom. 14 3Cell and Molecular Science, James Hutton Institute, Invergowrie, Dundee DD2 5DA, 15 United Kingdom. 16 aCurrent address: State Key Laboratory of Crop Biology, Shandong Provincial Key 17 Laboratory of Agricultural Microbiology, Shandong Agricultural University, Tai an, 18 Shandong, PR China. 19 bCurrent address: Plant Breeding division, Bangladesh Agricultural Research Institution. 20 1701-Gazipur, Bangladesh. 21 22 *Corresponding author: 23 e-mail: [email protected] 24 25 26 & These authors contributed equally to this work 27 28 29 30

1 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

31 Abstract 32 Potato late blight, which is caused by the destructive oomycete pathogen Phytophthora 33 infestans, is a major threat to global food security. Several nucleotide binding, leucine- 34 rich repeat (NLR) Resistance to P. infestans (Rpi) genes have been introgressed into 35 potato cultivars from wild Solanum species that are native to Mexico, but these were 36 quickly defeated. Positional cloning in Solanum mochiquense, combined with allele 37 mining in Solanum huancabambense, were used to identify a new family of Rpi genes 38 from Peruvian Solanum species. Rpi-mcq1, Rpi-hcb1.1 and Rpi-hcb1.2 confer race- 39 specific resistance to a panel of P. infestans isolates. Effector assays showed that the Rpi- 40 mcq1 family mediates a hypersensitive response upon recognition of the RXLR effector 41 AVR2, which had previously been found to be exclusively recognized by the family of 42 R2 resistance proteins. The Rpi-mcq1 and R2 genes are distinct and reside on 43 chromosome IX and IV, respectively. This is the first report of two unrelated R protein 44 families that recognize the same AVR protein. We anticipate that this likely is a 45 consequence of a geographically separated dynamic co-evolution of R gene families of 46 Solanum with an important effector gene of P. infestans. 47 48 Key words 49 Avr2, co-evolution, chromosome IX, Phytophthora, potato, resistance gene, Rpi-hcb1, 50 Rpi-mcq1. 51 52 Short title 53 Rpi-mcq1 resistance gene against Phytophthora infestans

54

2 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

55 Author summary 56 Potato is the largest non-grain staple crop and essential for food security world-wide. 57 However, potato plants are continuously threatened by the notorious oomycete pathogen 58 Phytophthora infestans that causes late blight. This devastating disease has led to the Irish 59 famine more then 150 years ago, and is still a major threat for potato. Resistance against 60 P. infestans can be found in wild relatives of potato, which carry resistance genes that 61 belong to the nucleotide binding site-leucine-rich repeat (NLR) class. Known NLR 62 proteins typically recognize a matching effector from Phytophthora, which leads to a 63 hypersensitive resistance response (HR). For example, R2 from Mexican Solanum 64 species recognizes AVR2 from P. infestans. So far, these R genes exclusively match to 65 one Avr gene. Here, we identified a new class of NLR proteins that are different from R2, 66 but also recognize the same effector AVR2. This new family of NLR occurs in South 67 American Solanum species, and we anticipate that it is likely a product of a 68 geographically separated co-evolution with AVR2. This is the first report of two unrelated 69 R protein families that recognize the same AVR protein. 70

3 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

71 Introduction 72 Potato (Solanum tuberosum L.) is the most important non-cereal crop directly consumed 73 worldwide and plays a pivotal role in food security. To date, the major threat for potato 74 production is the devastating late blight disease, which is caused by the Irish famine 75 pathogen Phytophthora infestans [1]. Breeding for resistance to late blight began in the 76 1850s, when high levels of resistance were found in wild Solanum demissum that is native 77 to the highlands of Toluca Valley in Mexico [2]. As a result of a tight co-evolution 78 between wild Solanum and P. infestans in this center of genetic diversity, a wealth of 79 resistance (R) genes have evolved in Mexican Solanum species [3-6]. Identified R genes 80 include R1-R11 from S. demissum, members from the Rpi-blb1, Rpi-blb2 and Rpi-blb3 81 family from S. bulbocastanum and S. stoloniferum, Rpi-amr3 from S. americanum, and 82 few more [7, 8]. Unfortunately, all R genes identified so far have failed to provide durable 83 resistance to the ‘R gene destroyer’ P. infestans [1]. A second center of genetic diversity 84 of P. infestans and tuber-bearing Solanum species occurs in South America [5]. More 85 recently, new R genes from those regions have been isolated as well, such as the Rpi-vnt1 86 family from Solanum venturii in Argentina and Rpi-ber family from Bolivian S. 87 chacoense, S. berthaultii and S. tarijense [9, 10]. In addition to these, various other South 88 American wild Solanum species are resistant to P. infestans and could provide useful R 89 genes against late blight [11]. 90 The oomycete pathogen P. infestans secretes an arsenal of effectors in order to manipulate 91 host defense responses. Cytoplasmic effectors are typically members of the RXLR class, 92 consisting of modular proteins with an N-terminal signal peptide, a conserved Arg-X- 93 Leu-Arg (RXLR) motif required for translocation inside of the host cells and a highly 94 polymorphic C-terminal domain [12]. Avr2 of P. infestans is a well characterized RXLR 95 effector and member of a highly diverse gene family. AVR2 targets the plant phosphatase 96 BSL1 that is implicated in positive regulation of the brassinosteroid (BR) signaling. BR- 97 signaling elevates expression of the transcription factor CHL1, which acts as a negative 98 regulator of immunity. AVR2 was found to play an important role promoting P. infestans 99 virulence and suppressing the effect of Pattern-Triggered immunity (PTI). Therefore, 100 Avr2 is considered an important effector gene and thus drives selection for R genes that 101 recognize it [13, 14]. 102 R2 resistance genes which mediate recognition of AVR2 have been cloned from a major 103 late blight resistance locus on the short arm of chromosome IV [15, 16]. Ten R2 homologs

4 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

104 were identified from the Mexican wild Solanum species S. bulbocastanum, S. demissum, 105 S. edinense, S. hjertingii, and S. schenckii [17, 18]. In addition to these, AVR2 was 106 recently found to be also recognized by Solanum mochiquense (mcq) and Solanum 107 huancabambense (hcb) that originate from Peru [18]. The activation of hypersensitive 108 cell death responses upon delivery of AVR2 suggests that additional R genes are present 109 in these Solanum species. 110 In this study, we focused on a new family of Rpi genes from the Peruvian S. mochiquense 111 and S. huancabambense. Rpi-mcq1, Rpi-hcb1.1 and Rpi-hcb1.2 were cloned by map- 112 based cloning in combination with a gene allele mining approach. The newly isolated Rpi 113 genes encode NLR proteins that are sequence-unrelated with the R2 family and confer 114 resistance to P. infestans isolates. We show here that diversifying selection is acting in 115 those genes, which may be a consequence of an independent molecular arms race between 116 wild populations of P. infestans and Solanum species from South America. 117 118 Materials and Methods 119 Plant materials 120 The interspecific Solanum mochiquense BC1 mapping population was developed by 121 crossing the P. infestans-resistant parent A988 (CGN18263) and susceptible parent A966 122 (CGN17731) [19]. The seeds were routinely treated with 1000 ppm gibberellic acid 123 (GA3) for 24 hours to break dormancy and were sown on MS20 medium. Recombinants 124 seedling plants were transferred to greenhouse and treated regularly with fungicides and 125 pesticides to control other pests as described previously by Foster, Park (20). 126 127 DNA isolation and sequencing 128 DNA isolation was performed using a Retch protocol [15]. BAC and cosmid clone DNA 129 were isolated using the Qiagen Midi Prep Kit ® (Qiagen). BACs and cosmids ends, PCR 130 products and shortgun clones were sequenced using the ABI PRISM BigDye® 131 Terminator v3.1 Cycle Sequencing kit (Applied Biosystens) using manufacturer 132 instructions. 133 134 PCR-based marker development 135 The PCR-based markers used in this study are listed in SI Table S1. Cleaved amplified 136 polymorphic sequence (CAPS) markers T0156 and TG328 were developed according

5 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

137 with Smilde, Brigneti (19). TG591N and S1D11 were developed from restriction 138 fragment length polymorphism (RFLP) markers annotated on Tomato-EXPEN map or 139 Potato Maps [21]. Blast searches with the tomato gene sequences were performed on the 140 Solanum tuberosum unigenes database at Sol genomics network [22] and the obtained 141 Expressed Sequenced Taq (EST) homologs were used for CAPS markers development. 142 U286446, U296361 and U272857 were developed from a released tomato BAC sequence 143 (HBa_165P17) which overlaps with the Rpi-mcq1 region. The CAPS marker 9C23R was 144 developed from the BAC end sequence of the candidate 9C23 BAC clone. PCR products 145 from resistance bulks and susceptible bulks were sequenced for each pair of primers. The 146 identified SNPs were used to develop additional CAPS markers which were mapped on 147 chromosome IX by using the 163 recombinants identified with T0156 and S1D11. 148 149 BAC library construction and screening 150 A BAC library was constructed with pIndigoBAC-5 (Hind III-Cloning Ready) for a 151 heterozygous resistant plant K182 carrying the Rpi-mcq1 followed the instrument 152 (Epicentre, WI, USA). The library is approximately 9× coverage with an average insert 153 size of 85 kb. The BAC library was pooled for each 384-well plate and screened by PCR 154 using the flanking and co-segregating makers. Positive clones were validated by singleton 155 PCR. 156 157 Allele mining 158 The Rpi-mcq1_RGH primers including the CACC site for Gateway® cloning purposes, 159 were designed on Rpi-mcq1 sequence (SI Table S1). Genomic DNA of the resistant 160 genotype hcb353-8 was used as template in a PCR reaction (98C: 30’'’, 25X [98C: 10’'’, 161 64C: 30’'’, 72C: 1’30’'] 72C; 10’'’) using Phusion® high-fidelity DNA polymerase 162 (Thermo Fisher Scientific). Amplicons were separated on agarose gel and purified using 163 the QIAquick Gel Extration Kit® (Qiagen). Purified products were cloned in pENTR/D- 164 TOPO® (Invitrogen) using DH10B E. coli competent cells. The inserts of the entry clones 165 were checked by PCR and Sanger sequencing. Unique sequences were transferred to the 166 pK7WG2 destination vector [23] by an LR reaction using LR-clonase II® (Invitrogen) 167 and plasmids were used for A. tumefaciens transformation strain AGL1. 168 169 Binary vectors for Agrobacterium tumefaciens-mediated transient transformation

6 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

170 Avr2, Avr3a, Rpi-hcb1.1 and Rpi-hcb1.2 were cloned in the binary vector pK7WG2 [23]. 171 R2, Rpi-mcq1 and R3a were previously cloned in the binary vector pBINPLUS [24]. The 172 NLR candidate genes cosA1, cosA2 and cosD5 were cloned in the binary cosmid vector 173 pCLD04541 [25]. 174 175 Agroinfiltration 176 Agroinfiltration was performed as described by Domazakis, Lin (26). Briefly, young and 177 fully expanded leaves of 4-5-week-old potato and N. benthamiana plants were used. A 178 suspension of A. tumefaciens strain AGL1 containing the appropriate expression vectors

179 at an OD600 of 0.2 for potato and at an OD600 of 0.5-1 for N. benthamiana were infiltrated 180 in leaf panels. Three plants per genotype and three leaves per plant were used in two 181 biological replicates. Local symptoms were assessed at 3-4 dpi. The percentage of cell 182 death was quantified using scores of 0%, 25%, 50%, 75% and 100%, based on 183 observation of the infiltrated area. 184 185 Transient complementation 186 Agroinfiltration of A. tumefaciens strains carrying Rpi-mcq1, Rpi-hcb1.1 or Rpi-hcb1.2 187 were performed in four-week-old N. benthamiana plants according with method 188 described by Domazakis, Lin (26) . Two days after inoculation, infiltrated leaves were 189 detached and inoculated with P. infestans isolates IPO-0 and IPO-C (S2 Table). Infection 190 symptoms were scored between 4-8 dpi. 191 192 Stable transformation 193 A. tumefaciens-mediated stable transformation was performed with potato cv. ‘Désirée’ 194 and tomato cv. ‘Moneymaker’. The cosmid vector pCLD04541 harboring the cosmids 195 A2, A1 and D5 under the control of their native promoters, were used in potato and tomato 196 transformation to test the function of the NLR candidate genes. pK7WG2:Rpi-hcb1.1 and 197 pK7WG2:Rpi-hcb1.2 under the control of 35S promoter were used in potato 198 transformation to test the function of Rpi-hcb genes. The transformation was performed 199 using routine transformation protocols [27]. Transformants were selected after growth in 200 greenhouse conditions (18-22°C, 16 h of light and 8 h of dark). 201 202 Phytophthora infestans isolates, culture conditions and inoculum preparation

7 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

203 The P. infestans isolates used in this study are listed in S2 Table. Isolates were retrieved 204 from our in-house stock collection. Isolates were grown in the dark at 15°C on solid rye 205 sucrose medium prior to the disease test [28]. To isolate zoospores for plant inoculations, 206 sporulation mycelium was flooded with cold water and incubated at 4°C for 1-3 hours. 207 208 Disease test 209 Leaves from 6-8-week-old plants grown in greenhouse conditions (18-22°C, 16 h of light 210 and 8 h of dark) were detached and placed in water-saturated oasis in trays. The leaves 211 were spot-inoculated at the abaxial leaf side with 10µl droplets containing 5*104 212 zoospores per ml (in tap water). After inoculation, the trays were incubated in a climate 213 chamber at 15°C with a 16h photoperiod. Development of lesions and presence of 214 sporulation was determined at 5-6 dpi. Disease index was estimated using a scale from 1 215 to 9, ranging from expanding lesions with massive sporulation (1 to 3, susceptible), 216 sporulation no clearly visible (4), sporadic sporulation only visible under the microscope 217 (5), lesion with a diameter size ≥ 10 mm (6), occurrence of hypersensitive response (HR) 218 between 3 to 10 mm (7), less abundant sporulation and smaller lesions, occurrence of HR 219 (8, resistant) and to no symptoms (9, fully resistant). 220 In all the cases, three to five leaves were inoculated per isolate, with 12 inoculation spots 221 per leaf for potato (Désirée) and tomato (Moneymaker) transformants and 3 spots for N. 222 benthamiana experiments. At least three independent experiments were performed. 223 224 Phylogenetic and positive selection analysis 225 A UPGMA-based tree was generated with the full amino acid sequence of 16 proteins, 226 including Rpi-mcq1 and R2 family members. Bootstrap value was set equal to 100. The 227 obtained UPGMA-based tree was displayed as circular unrooted cladogram using 228 Geneious®9.1.2. 229 A codon-based analysis was conducted using PAMLX1.3.1 package [29]. Maximum- 230 likelihood codon substitution models M0, M1, M2, M7 and M8 were used for the 231 analysis. Positively selected sites detected by, Models M2 and M8 were identified using 232 Empirical Bayes Statistics [30]. 233 234 Results 235 High resolution mapping and cloning of Rpi-mcq1

8 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

236 Rpi-mcq1 (formerly named Rpi-moc1) was previously mapped to a region of 15.8 cM on 237 chromosome IX, flanked by the CAPS marker T0156 and an AFLP marker [19]. In order 238 to fine map Rpi-mcq1, a BC1 mapping population from S. mochiquense accessions A988 239 and A966 was developed. A recombinant screen on 2502 individuals from the mapping 240 population using the flanking markers S1D11 and T0156 was performed. 163 241 recombinant individuals were identified and characterized for resistance to P. infestans 242 isolates 90128 and EC1, resulting in resistant and susceptible bulks. Subsequently, five 243 CAPS markers including TG328, U286446, U296361, TG591N and U272857 were 244 screened on the recombinants (S1 Table). Rpi-mcq1 was mapped to a narrow region 245 flanked by U286446 and S1D11, and co-segregates with TG591N and U282757 (S1 Fig). 246 U286446, TG591N and U282757 were tested on a BAC library, and two overlapping 247 BAC clones 9C23 and 43B09 derived from the resistant haplotype were identified (Fig 248 1a). A new CAPS marker, 9C23R, was obtained from the end sequence of BAC clone 249 9C23 and was screened, revealing segregation of Rpi-mcq1 with six recombinants in the 250 bulks. The marker 9C23R is located on the opposite orientation of Rpi-mcq1 at a distance 251 of 0.24 cM (Fig 1a, S1 Fig). Rpi-mcq1 was fine-mapped on the two overlapping BACs, 252 9C23 and 43B09. Subsequently, the two BACs were sequenced by TIGR and as a result 253 two contigs that shared 62,395 and 114,083 bp, respectively, were obtained. In total eight 254 ORFs were predicted in the two contigs, including four Tm22 - like NLR candidate genes 255 and four non-NLRs, including a putative NAD dependent epimerase, a RNA-directed 256 DNA polymerase, a retransposon, and a protein with unknown function. Two NLR 257 candidate genes, cosA2 and cosA1, have full length ORFs with ~80-85% identity to 258 Tm22 at the amino acid level. A third NLR candidate gene, cosD5, represents a partial 259 NLR with a truncated Coil-Coil (CC) domain. The fourth NLR candidate gene, cosE7, 260 contains an early stop codon and was omitted from further study. 261 262 Rpi-mcq1 confers late blight resistance in potato and tomato 263 To determine whether the genes cosA2, cosA1 and cosD5 confer resistance to P. infestans 264 in potato and tomato, stable potato and tomato transformants cv. ‘Désirée’ and cv. 265 ‘Moneymaker’, respectively, were generated. In total, 22, 24, and 20 independent primary 266 transformants were obtained in potato that express the cosA2, cosA1 and cosD5 genes, 267 respectively, and 8, 7, and 10 independent tomato transgenic lines were obtained. 268 Following transfer to the greenhouse, a detached leaf assay was performed on putative 269 transformants and in the wild type ‘Désirée’. Detached leaves of the selected lines were

9 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

270 spot-inoculated with the avirulent P. infestans isolates EC1 and 90128 (S2 Table). 271 Macroscopic observations were carried out at 6 days post inoculation (dpi). On potato, 272 13 out of 22 putative transformants derived from cosA2 showed hypersensitive response 273 (HR) and were resistant to the tested P. infestans isolates, whereas abundant sporulation 274 was found in the 44 putative transformants derived from cosA1 and cosD5 and in the wild 275 type ‘Désirée’ (Fig 1b). Upon inoculation with the virulent P. infestans isolate IPO-C (S2 276 Table), all plants were infected (Fig 1b). Consistent with these results, only the putative 277 transformants derived from cosA2 showed race-specific resistance to EC1 and 90128 in 278 tomato cv. ’Moneymaker' (S2 Fig). Altogether, the results indicate that cosA2 can 279 complement the late blight susceptible phenotype in potato and tomato, and we 280 designated it Rpi-mcq1. 281 282 Rpi-mcq1 homologs are detected in Solanum huancabambense 283 The wild S. mochiquense that carries Rpi-mcq1 was previously shown to mount AVR2- 284 specific cell death upon agroinfiltration [18]. In high-throughput effector screens in 285 resistant wild Solanum germplasm, also S. huancabambense (hcb) was found to show 286 AVR2-triggered cell death. To confirm the specificity of this response in independent 287 experiments, Avr2 was transiently expressed by agroinfiltration in hcb353-8 and in the 288 wild type ‘Désirée’. Single infiltrations of empty vector pK7WG2 and co- 289 agroinfiltrations of R3a/Avr3a were included as negative and positive controls, 290 respectively. Specific cell death responses to AVR2 were evident in hcb353-8 and no cell 291 death responses were identified on ‘Désirée’, suggesting the presence of an AVR2- 292 recognizing R protein in hcb353-8 (S3a Fig). 293 To determine the resistance spectrum of hcb353-8, a detached leaf assay was performed 294 with a panel of 12 P. infestans isolates. From this set, four and eight isolates are virulent 295 or avirulent on R2 plants, respectively (S2 Table). Among the twelve isolates, resistance 296 phenotypes on hcb353-8 were fully correlating with R2-specific resistance (S2 Table). At 297 6 dpi, hcb353-8 showed a typical HR and was resistant to the avirulent isolate IPO-0, 298 while large sporulating lesions were found in the cv. ‘Désirée' (S3b Fig). These results 299 indicate that a race-specific Rpi gene is present in hcb353-8, which exhibits similarities 300 to AVR2-based resistance. 301 To investigate whether the Rpi gene in hcb353-8 is homologous to Rpi-mcq1, a genetic 302 analysis was performed. The resistant hcb353-8 was crossed with the susceptible hcb354- 303 2. Detached leaves of the parents and the 18 offspring genotypes were inoculated with P.

10 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

304 infestans isolates 90128 and IPO-0. Macroscopic observations were carried out at 6 dpi. 305 The susceptible hcb354-2 was infected by both isolates, whereas hcb353-8 showed HR 306 and was resistant. No disease symptoms were found in ten genotypes of the progeny, 307 whereas eight genotypes showed abundant sporulation and disease symptoms. This result 308 confirms an expected 1:1 segregation ratio (x2= 0.11, P=0.74), suggesting the presence of 309 a single dominant Rpi gene in hcb353-8 (S3a Table). Subsequently, to determine the 310 genetic position of the Rpi in hcb353-8, a set of PCR markers that reside on LG IX of 311 potato were tested on the F1 progeny (S1 Table). The marker GP41 revealed 312 polymorphism between the parents and progeny of the population and a complete co- 313 segregation of marker patterns with P. infestans resistance was observed (S3b Table). 314 Additionally, we agroinfiltrated the progeny plants with Avr2 and we found that the 315 response to AVR2 segregates with resistance to P. infestans (S3c Table). In summary, 316 the Rpi in hcb353-8 is located on LG IX and together with the AVR2-based race-specific 317 resistance, we hypothesize that the Rpi gene in hcb353-8 is homologous to Rpi-mcq1. 318 We followed a homology-based cloning approach to identify the Rpi gene. PCR with the 319 conserved Rpi-mcq1 primers (Rpimcq1_RGH) (S1 Table) on genomic DNA of hcb353-8 320 plants yielded amplicons of 2577 bp, which is the range of the expected size of an Rpi- 321 mcq1 homologue. Sequence analysis of the amplicons revealed five additional Rpi-mcq1 322 gene homologues (RGH) from hcb353-8. Amino acid similarities among these 323 homologues vary between 73.3% to 86.6%, however, the similarity to members of the R2 324 family is much lower, namely between 42.4% to 45.6% (Fig 2, S4 Table). The 325 phylogenetic relationship between Rpi-mcq1, RGH and the R2 homolog Rpi-blb3 was 326 examined. A neighbor joining (NJ) tree grouped Rpi-mcq1 and the RGH in a single clade, 327 separate from Rpi-blb3 (S4 Fig). This result confirms that the Rpi-blb3/R2 and Rpi-mcq1 328 families are sequence unrelated and represent separate clades. 329 330 Rpi-hcb1.1 and Rpi-hcb1.2 respond to AVR2 and confer resistance to P. infestans 331 To test the function of the RGH, the five identified homologues were cloned in binary 332 vector pK7WG2. Rpi-mcq1 was previously cloned in the binary vector pBINPLUS [18]. 333 The constructs were transferred to A. tumefaciens strain AGL1. Co-agroinfiltrations were 334 performed in potato cv. ‘Bintje’ leaves with A. tumefaciens expressing each RGH or Rpi- 335 mcq1 with Avr2. Single infiltrations of Avr2, RGH or Rpi-mcq1 and empty vector were 336 included as negative controls. Co-agroinfiltration of R3a/Avr3a was included as positive

11 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

337 control. Two RGH, i.e. RGH4 and RGH5 induced specific cell death with AVR2 and we 338 designated them Rpi-hcb1.1 and Rpi.hcb1.2, respectively (Figs 2 and 3). 339 For assessing whether the Rpi-hcb genes confer resistance to P. infestans, a transient 340 complementation assay was conducted. Rpi-mcq1, Rpi-hcb1.1 and Rpi-hcb1.2 were 341 agroinfiltrated in N. benthamiana leaves. Additionally, agroinfiltrations with Rpi-blb3, 342 empty vector, and infiltration medium (MMA) were used as a positive and negative 343 controls, respectively. Two days later, zoospore suspensions of the avirulent IPO-0 and 344 virulent IPO-C P. infestans isolates (S2 Table) were spot-inoculated on the agroinfiltrated 345 leaf panels [18]. After 8 days, Rpi-mcq1, Rpi-hcb1.1, Rpi-hcb1.2, and Rpi-blb3-treated 346 leaf panels display a HR to IPO-0 (race 0), whereas large expanding necrotic lesions 347 surrounded by sporulation zone were observed on leaves with the complex race IPO-C. 348 Leaf panels agroinfiltrated with the empty vector and MMA medium show large 349 sporulating lesions for both isolates (S5 Fig). These data show that Rpi-hcb1.1 and Rpi- 350 hcb1.2 confer a race-specific resistance to P. infestans, similar to Rpi-mcq1- and Rpi- 351 blb3-mediated resistance. 352 To confirm the results with the transient complementation assay in potato, stable potato 353 transformants cv. ‘Désirée’ were generated that express the Rpi-hcb genes under the 354 control of the 35S constitutive promoter. 18 Rpi-hcb1.1 and 24 Rpi-hcb1.2 independent 355 primary transformants were selected and cultured in greenhouse. Subsequently, they were 356 tested for AVR2 response by agroinfiltration. Four and five independent transgenic lines 357 expressing Rpi-hcb1.1 or Rpi-hcb1.2, respectively, were selected. Detached leaves of the 358 selected lines were inoculated with two avirulent P. infestans isolates PIC99183 and IPO- 359 0 (S2 Table). Stable transformed plants containing the R genes were resistant to both 360 tested isolates. This result confirms that Rpi-hcb1.1 as well as Rpi-hcb1.2 confer 361 resistance to P. infestans in potato (Fig 4). 362 363 Rpi-mcq1 is under diversifying selection 364 The Rpi-mcq1 family harbors all characteristics of NLR genes, and overall, the six Rpi- 365 mcq1 family members display considerable sequence conservation in the CC and NB- 366 ARC domain. In contrast, the LRR domain, that is typically involved in effector 367 recognition, contains multiple single nucleotide polymorphisms (SNPs). To get further 368 insights on the region of Rpi-mcq1 that is under diversifying selection, the PAML method 369 was conducted [30]. Several positive selected amino acid residues were identified on the 370 CC, the NB-ARC and the LRR domain (S6 Fig). Model M2 identified 11 amino acid

12 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

371 residues under positive selection, from which 10 amino acids were present in the LRR 372 domain and 1 in the NB-ARC domain. The selection model M8 predicted the same 11 373 amino acids identified in M2 model and 6 additional amino acids. 1 out of 6 was present 374 in CC domain, and the other amino acids were present in the LRR domain. This points to 375 a strong diversifying selection on the LRR domain (S6 Fig). This suggests that the 376 polymorphisms could be driven by the evolution of differential specificities of AVR 377 recognition that is typically located in the LRR domain [31]. 378 379 Discussion 380 Despite cloning of a number of Rpi genes in the past years [10], the fast-evolving P. 381 infestans remains the most threatening pathogen of potato worldwide. To avoid losing the 382 arms-race with this fast-evolving oomycete pathogen, it is necessary to intensify the 383 search for new sources of resistance and extend the repertoire of available Rpi genes for 384 breeding. The majority of introgressed resistance genes into potato cultivars, have been 385 cloned from Solanum species native to Mexico, however, South America that is 386 considered the second center of genetic diversity has been much less explored. In this 387 manuscript, we report the cloning and functional characterization of a new family of late 388 blight resistance genes from the Peruvian S. mochiquense and S. huancabambense. Rpi- 389 mcq1, Rpi-hcb1.1 and Rpi-hcb1.2 confer race-specific resistance to P. infestans and 390 belong to the Rpi-mcq1 locus on linkage group (LG) IX [19]. The three Rpi-mcq1 391 homologs share high levels of amino acid sequence identity, and we found that all three 392 Rpi proteins mediate response to AVR2 of P. infestans. 393 Known Avr genes of P. infestans typically seem to have co-evolved with a single R locus 394 in potato, such as R3a and Avr3a [32, 33]. However, AVR2 is the first example of an 395 AVR protein that is recognized by two different, sequence-unrelated R protein families 396 in tuber-bearing Solanum species. In Mexican Solanum species, AVR2 recognition is 397 conferred by R2 family on chromosome IV, whilst in Peruvian Solanum genotypes AVR2 398 response is mediated by Rpi-mcq/hcb family located on chromosome IX [18]. Since Avr2 399 has a strong virulence function and is broadly spread among the pathogen strains, multiple 400 R gene families seem to have evolved to recognize this effector [13, 14]. The R gene 401 families display dissimilar sequences, the Solanum species occur in different 402 phylogenetic clades and the plants occur in different geographic regions, indicating that 403 they have evolved independently [18]. Our evolutionary analyses with Rpi-mcq/hcb

13 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

404 sequences indicate that the LRR motif, which plays a pivotal role in effector recognition 405 specificity, is shaped by a strong diversifying selection that could result in new 406 specificities of effector recognition [17, 34]. Similar properties were found for the R2 407 family [17]. These findings are in line with the concept of fast nucleotide evolution and 408 sequence interchange between R gene homologs, as major mechanism shaping R gene 409 diversity in plants, and especially targeted at the LRR [35, 36]. 410 The recognition of an AVR protein by different R proteins has been reported for other 411 plant-pathosystems, however, the relationship of those R genes has so far been unknown. 412 For instance, Avr3a/5 of Phytophthora sojae is recognized by the soybean resistance 413 genes Rps3 as well as Rps5 from which the sequence is still unreported [37]. Avr-Rmg7/8 414 of Magnaporthe grisea is recognized by the wheat resistance genes Rmg7 and Rmg8, 415 which are located in homeologous chromosomes 2B and 2A, respectively [38]. 416 Furthermore, some rice R genes were recently found to interact with divergent 417 Magnaporthe oryzae effectors via different binding surfaces [39, 40]. The effectors have 418 evolved independently to unconventional R gene domains and target them with different 419 binding-specificities [41]. These studies are providing more insights in the antagonistic 420 interplay between pathogen and host that is driven by co-evolutionary forces targeted at 421 R and Avr genes. 422 Pathogen populations in European countries, the USA and Canada have developed 423 virulence to potato plants carrying the S. demissum-derived R1, R3, R4, R7, R10 and R11. 424 However, lower frequencies were noted for virulence on R2, and R2 significantly delays 425 the onset of epidemics in field trails [42, 43]. Since these characteristics heavily rely on 426 the matching Avr gene, also Rpi-mcq/hcb are expected to display extended latency periods 427 in the field and contribute to late blight resistance in practice. 428 Our improved understanding of the recognition of Avr2 by R2 and Rpi-mcq/hcb confirms 429 that effectors are a powerful tool to accelerate the identification of R genes. This study is 430 sharpening the concept that effector-based breeding should be complemented with 431 additional molecular techniques, e.g molecular markers, to securely determine the 432 identity of the matching R genes. Combination of Rpi genes from distinct loci but with 433 similar race-specific resistance specificities, such as R2 and Rpi-mcq1, could be more 434 effective than using single R genes, however, a broader resistance spectrum is not 435 expected based on AVR2 recognition alone. However, Avr2 is member of a highly diverse 436 gene family, and molecular studies on Avr2 variants will reveal deeper insights 437 underlying recognition specificities and pathogenicity in P. infestans populations. Such

14 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

438 knowledge should contribute to wiser strategies for efficient deployment of R genes in 439 potato. 440 441 Acknowledgements 442 This work was supported by NWO-VIDI grant 12378 (V.G.A.A.V), COLCIENCIAS 443 doctoral grant 617-2013 (C.A-G), The Veenhuizen Tulp Fund (C.A-G), COST action 444 FA1208 (V.G.A.A.V, C.A-G, E.M.G, P.B., J.D.G.J). We thank Gert van Arkel for 445 technical assistance, Isolde Pereira for plant maintenance, Geert Kessel, Francine Govers 446 and David Cooke for providing P. infestans strains. Dionne Turnbull and Susan Breen 447 are acknowledged for assistance in preliminary co-agroinfiltrations. 448 449 Figure legends 450 Fig 1 Physical map of the S. mochiquense genomic region containing Rpi-mcq1 and 451 genetic complementation. (a) Black rectangles represent bacterial artificial chromosome 452 (BAC) clones (9C23 and 43B09) covering the region of Rpi-mcq1 and cosmid clones 453 carrying the NLR candidate genes. Blue lines indicate two contigs sequences from two 454 BAC clones. Colored segments represent four NLR candidate genes, two of them encoded 455 a complete CC-NB-LRR genes, represented as cosA2 and cosA1, respectively. Asterisk 456 indicates an early stop codon in one NLR candidate gene. (b) Representative pictures of 457 the disease symptoms of potato transgenic plants inoculated with P. infestans isolates 458 90128, EC1 and IPO-C, respectively. Pictures were taken at 6 dpi. PL2796 and PL2804 459 represent two transgenic lines obtained from cosA2 construct, PL2768 and PL2816 460 represent transgenic lines obtained with cosA1 and cosD5 constructs, respectively. 461 462 Fig 2 Protein alignment of the Rpi-mcq1 family. Multiple sequence alignment of Rpi- 463 mcq1, Rpi-hcb1.1 and Rpi-hcb1.2, using the Rpi-mcq1 amino acid sequence as reference. 464 The CC, NB-ARC and LRR domains are shaded in green, blue and grey, respectively. 465 The characteristics conserved domains including the P-loop (Kinase 1a), Kinase 2, kinase 466 3 and the RNBS-D (resistance domain) in the NB-ARC are underlined. The 14 LRR 467 repeats and the hydrophobic amino acids are represented with black rectangles and bold 468 letters, respectively. 469

15 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

470 Fig 3 Rpi-mcq1 homologues mediate response to AVR2. Agrobacterium-mediated 471 expression of Avr2 with Rpi-mcq1, Rpi-hcb1.1 and Rpi-hcb1.2 in potato cv. ‘Bintje’. 472 Single infiltrations of each R gene alone, Avr2 and empty vector were included as 473 negative controls, and co-agroinfiltrations of R3a/Avr3a as positive control. 474 475 Fig 4 Rpi-hcb1.1 and Rpi-hcb1.2 confer late blight resistance. Four and five transgenic 476 Désirée lines expressing Rpi-hcb1.1 or Rpi-hcb1.2, respectively, were inoculated with P. 477 infestans isolates PIC99183 and IPO-O. (a) Representative photographs are shown for 478 Désirée expressing Rpi-hcb1.1 lines 24 and 27 and Rpi-hcb1.2 lines 22 and 52, 479 respectively, at six dpi. Wild type (wt) ‘Désirée’ represents a susceptible control. (b) 480 Disease symptoms on ‘Désirée’ (WT), Désirée-Rpi-hcb1.1 lines 7, 22, 24 and 27 and 481 Désirée-Rpi-hcb1.2 lines 2, 22, 35, 52 and 56, are scored on a scale from 1 (susceptible) 482 to 9 (resistant). 483 484 Supporting information 485 S1 Fig. High-resolution genetic linkage map of the Rpi-mcq1 locus on chromosome 486 IX. 487 Genetic distances and marker names are indicated on the left and right of the map, 488 respectively. Rpi-mcq1 co-segregates with polymorphism makers TG591N and U272857, 489 flanked by 9C23R and U286446 (or U296361), respectively. 490 491 S2 Fig. Genetic complementation for late blight resistance on tomato with NLR 492 candidate genes. Representative pictures of the disease symptoms of PL3320, PL2741 493 and PL2759 tomato transgenic lines expressing the NLR candidate genes cosA2, cosA1 494 and cosD5, respectively. Detached leaves were inoculated with P. infestans isolates 495 90128 and EC1 and macroscopic observations for disease symptoms were carried out at 496 6 dpi. 497 498 S3 Fig. Solanum huancabambense recognizes AVR2 and is resistant to P. infestans. 499 (a) Agroinfiltration of A. tumefaciens carrying pK7WG2:Avr2 and co-infiltrations of 500 R3a/Avr3a on hcb353-8 and the wild type ‘Désirée. Single infiltrations of the pK7WG2 501 empty vector and co-infiltrations of R3a/Avr3a and/or R2/Avr2 were included as negative 502 and positive controls, respectively. (b) hcb353-8 shows no lesion development upon

16 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

503 inoculation with P. infestans isolate IPO-O, whereas ‘Désirée’ (wt) shows large 504 sporulating lesions at 6 dpi. Representative pictures are presented. 505 506 S4 Fig. Phylogenetic relationship and sequence similarity of Rpimcq1 family. The 507 UPGMA-based tree illustrates the phylogenetic relationship at the amino acid level of 508 Rpi-mcq1 and RGH cloned from hcb353-8 (blue) and the R2 family (red). Numbers on 509 each node represent bootstrap values based on 100 replicates. 510 511 S5 Fig. Genetic complementation of Rpi-mcq1, Rpi-hcb1.1 and Rpi-hcb1.2 in N. 512 benthamiana. Rpi-mcq1, Rpi-hcb1.1 and Rpi-hcb1.2 were transiently expressed in N. 513 benthamiana leaves, followed by inoculation with the P. infestans isolates IPO-0 and 514 IPO-C. Rpi-blb3 was included as a resistant control and infiltrations with medium (MMA) 515 and with empty vector were used as negative controls. Race-specific resistance to IPO-O 516 was observed with Rpi-mcq1, Rpi-hcb1.1, Rpi-hcb1.2 and in the resistant control Rpi- 517 blb3, whereas IPO-C caused expanding lesions. 518 519 S6 Fig. Positive selection in Rpi-mcq1 variants has mostly targeted the LRR domain. 520 (a) Graphical representation of posterior probability of diversifying selection based on 521 model M8 at each site of Rpi-mcq1 variants. The * indicates a posterior probability higher 522 than 99% of having 휔>1. The x axis denotes codon position in the alignment of Rpi-mcq1 523 variants made from codeml and removing all the gaps. (b) * Log likelihood value. **

524 Likelihood radio test: = 2(lnlalternative hypothesis- lnlnull hypothesis), with significance evaluated 525 from distribution: df is degree of freedom and p is the probability. *** Bayes Empirical 526 Bayer (BEB) analysis; amino acid sites, based on Rpi-hcb1.1 sequence, inferred to be 527 under diversifying selection with probability >95%, and 99% in bold. 528 529 S1 Table. Overview of CAPS markers and primers used in this study. Primer 530 sequence, annealing temperature and restriction enzyme are indicated. a Orientation of 531 the primer: F, Forward; R, reverse. b Annealing temperature. C Restriction enzyme that 532 reveals polymorphism between alleles linked to resistance or susceptibility. 533 534 S2 Table. List of P. infestans isolates used in this study.

17 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

535 The country, year, the genotype source of collection, the race, as well as the virulence (V) 536 and avirulence (A) on S. huancabambense (hcb) 353-8 and S. hjertjingii (hjt) 349-3 plants 537 are indicated. Hjt349-3 contains an R2 homolog [18] 538 539 S3 Table. Co-segregation for resistance, AVR2 response and genetic marker in hcb 540 7393 population. (a) Resistance (R), susceptibility (S) or quantitative resistance (Q) 541 against P. infestans isolates IPO-0 and 90128. (b) Presence (1) or absence (0) of the 542 polymorphic band of the genetic marker GP41on chromosome IX. (c) Presence (+) or 543 absence (-) of cell death response after agroinfiltration with Avr2, empty vector (negative 544 control) and co-infiltration with R3a/Avr3a (positive control) at 4 dpi. n.d. not 545 determined. 546 547 S4 Table. Sequence similarities between Rpi-mcq1/hcb and R2 family. Percentages 548 of amino acid sequence similarities between ten R2 family members and five identified 549 resistance gene homologues (RGH) of Rpi-mcq1 are presented.

550

18 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

551 References 552 1. Fry W. Phytophthora infestans: the plant (and R gene) destroyer. Molecular Plant 553 Pathology. 2008;9(3):385-402. doi: 10.1111/j.1364-3703.2007.00465.x. 554 2. Muller KO, Black W. Potato breeding for resistance to blight and virus diseases 555 during the last hundred years. Zeitschrift für Pflanzenzüchtung 1952;31:305-18. 556 3. Grunwald NJ, Flier WG. The biology of Phytophthora infestans at its center of 557 origin. Annual Review of Phytopathology. 2005;43:171-90. 558 4. Hawkes JG. The potato: evolution, biodiversity, and genetic resources. United 559 Kindom: Belhaven Press; 1990. 259 p. 560 5. Hijmans RJ, Spooner DM. Geographic distribution of wild potato species. 561 American journal of botany. 2001;88(11):2101-12. 562 6. Fry W, Goodwin SB, Dyer AT, Matuszak JM, Drenth A, Tooley PW, et al. 563 Historical and recent migrations of Phytophthora infestans: chronology, pathways, and 564 implications. Plant Disease. 1993;77:635- 61. 565 7. Witek K, Jupe F, Witek AI, Baker D, Clark MD, Jones JDG. Accelerated cloning 566 of a potato late blight-resistance gene using RenSeq and SMRT sequencing. Nature 567 Biotechnology. 2016;34:656-60. 568 8. Vleeshouwers VGAA, Raffaele S, Vossen JH, Champouret N, Oliva R, Segretin 569 ME, et al. Understanding and exploiting late blight resistance in the age of effectors. 570 Annual Review Phytopathology. 2011;49:507-31. doi: 10.1146/annurev-phyto-072910- 571 095326. 572 9. Sliwka J, Jakuczun H, Chmielarz M, Hara-Skrzypiec A, Tomczynska I, Kilian A, 573 et al. Late blight resistance gene from Solanum ruiz-ceballosii is located on potato 574 chromosome X and linked to violet flower colour. BMC Genetics. 2012;13(11):1-12. 575 10. Jo K-R, Zhu S, Bai Y, Hutten RCB, Kessel GJT, Vleeshouwers VGAA, et al. 576 Problematic Crops: 1. Potatoes: Towards sustainable potato late blight resistance by 577 cisgenic R gene pyramiding. Plant Pathogen Resistance Biotechnology: John Wiley & 578 Sons, Inc; 2016. p. 171-91. 579 11. Vleeshouwers VGAA, Finkers R, Budding D, Visser M, Jacobs MM, van Berloo 580 R, et al. SolRgene: an online database to explore disease resistance genes in tuber-bearing 581 Solanum species. BMC Plant Biology. 2011;11(1):1-9. doi: 10.1186/1471-2229-11-116. 582 12. Whisson SC, Boevink PC, Moleleki L, Avrova AO, Morales JG, Gilroy EM, et 583 al. A translocation signal for delivery of oomycete effector proteins into host plant cells.

19 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

584 Nature. 2007;450(7166):115-8. Epub 2007/10/05. doi: 10.1038/nature06203. PubMed 585 PMID: 17914356. 586 13. Turnbull D, Yang L, Naqvi S, Breen S, Welsh L, Stephens J, et al. RXLR effector 587 AVR2 up-regulates a brassinosteroid-responsive bHLH transcription factor to suppress 588 immunity. Plant Physiology. 2017;174(1):356-69. Epub 2017/03/09. doi: 589 10.1104/pp.16.01804. PubMed PMID: 28270626; PubMed Central PMCID: 590 PMCPMC5411136. 591 14. Gilroy EM, Breen S, Whisson SC, Squires J, Hein I, Kaczmarek M, et al. 592 Presence/absence, differential expression and sequence polymorphisms between PiAVR2 593 and PiAVR2-like in Phytophthora infestans determine virulence on R2 plants. New 594 Phytologist. 2011;191(3):763-76. doi: 10.1111/j.1469-8137.2011.03736.x. 595 15. Park TH, Gros J, Sikkema A, Vleeshouwers VG, Muskens M, Allefs S, et al. The 596 late blight resistance locus Rpi-blb3 from Solanum bulbocastanum belongs to a major late 597 blight R gene cluster on chromosome 4 of potato. Molecular Plant- Microbe Interaction. 598 2005;18(7):722-9. doi: 10.1094/mpmi-18-0722. 599 16. Lokossou AA, Park T-h, van Arkel G, Arens M, Ruyter-Spira C, Morales J, et al. 600 Exploiting knowledge of R/Avr genes to rapidly clone a new LZ-NBS-LRR family of late 601 blight resistance genes from potato linkage group IV. Molecular Plant-Microbe 602 Interactions. 2009;22(6):630-41. doi: 10.1094/mpmi-22-6-0630. 603 17. Champouret N. Functional genomics of Phytophthora infestans effectors and 604 Solanum resistance genes [Dissertation]. The Netherlands: Wageningen University; 605 2010. 606 18. Aguilera-Galvez C, Champouret N, Rietman H, Lin X, Wouters D, Chu Z, et al. 607 Two different R gene Loci co-evolved with Avr2 of Phytophthora infestans and confer 608 distinct resistance specificities in potato. Studies in Mycology. 2018;89:1-11. doi: doi: 609 10.1016/j.simyco.2018.01.002. 610 19. Smilde WD, Brigneti G, Jagger L, Perkins S, Jones JD. Solanum mochiquense 611 chromosome IX carries a novel late blight resistance gene Rpi-moc1. Theoretical Applied 612 Genetics. 2005;110(2):252-8. Epub 2005/01/27. doi: 10.1007/s00122-004-1820-8. 613 PubMed PMID: 15672258. 614 20. Foster SJ, Park TH, Pel M, Brigneti G, Śliwka J, Jagger L, et al. Rpi-vnt1.1, a Tm- 615 22 homolog from Solanum venturii, confers resistance to potato late blight. Molecular 616 Plant- Microbe Interactions. 2009;22(5):589-600. doi: 10.1094/mpmi-22-5-0589.

20 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

617 21. Riaño-Pachón DM, Nagel A, Neigenfind J, Wagner R, Basekow R, Weber E, et 618 al. GabiPD: the GABI primary database—a plant integrative ‘omics’ database. Nucleic 619 Acids Research. 2009;37(suppl_1):D954-D9. doi: 10.1093/nar/gkn611. 620 22. Fernandez-Pozo N, Menda N, Edwards JD, Saha S, Tecle IY, Strickler SR, et al. 621 The Sol Genomics Network (SGN)—from genotype to phenotype to breeding. Nucleic 622 Acids Research. 2015;43(D1):D1036-D41. doi: 10.1093/nar/gku1195. 623 23. Karimi M, Inzé D, Depicker A. GATEWAY vectors for Agrobacterium-mediated 624 plant transformation. Trends in Plant Science. 2002;7(5):193-5. 625 24. van Engelen FA, Molthoff JW, Conner AJ, Nap JP, Pereira A, Stiekema WJ. 626 pBINPLUS: an improved plant transformation vector based on pBIN19. Transgenic 627 research. 1995;4(4):288-90. Epub 1995/07/01. PubMed PMID: 7655517. 628 25. Tao Q, Zhang HB. Cloning and stable maintenance of DNA fragments over 300 629 kb in Escherichia coli with conventional plasmid-based vectors. Nucleic acids research. 630 1998;26(21):4901-9. PubMed PMID: 9776751. 631 26. Domazakis E, Lin X, Aguilera-Galvez C, Wouters D, Bijsterbosch G, Wolters PJ, 632 et al. Effectoromics-based identification of cell surface receptors in potato. In: Shan L, 633 He P, editors. Plant Pattern Recognition Receptors: Methods and Protocols. New York, 634 NY: Springer New York; 2017. p. 337-53. 635 27. Fillatti JJ, Kiser J, Rose R, Comai L. Efficient Transfer of a Glyphosate Tolerance 636 Gene into Tomato Using a Binary Agrobacterium Tumefaciens Vector. Bio/Technology. 637 1987;5(7):726-30. doi: 10.1038/nbt0787-726. PubMed PMID: 638 ScienceOpenVid:5b97be83-e42e-4a19-84c1-1ad849ba2512. 639 28. Caten CE, Jinks JL. Spontaneous variability of single isolates of Phytophthora 640 infestans. I Cultural variation. Canadian Journal of Botany. 1968;46(4):329-48. 641 29. Xu B, Yang Z. PAMLX: A graphical user interface for PAML. Molecular biology 642 and evolution. 2013;30(12):2723-4. doi: https://doi.org/10.1093/molbev/mst179. 643 30. Yang Z, Wong WSW, Nielsen R. Bayes empirical bayes inference of amino acid 644 sites under positive selection. Molecular biology and evolution. 2005;22(4):1107- 18. doi: 645 https://doi.org/10.1093/molbev/msi097. 646 31. Moffett P. Chapter 1 - Mechanisms of recognition in Dominant R Gene mediated 647 resistance. In: Loebenstein G, Carr JP, editors. Advances in Virus Research. 75: 648 Academic Press; 2009. p. 1-229. 649 32. Armstrong MR, Whisson SC, Pritchard L, Bos JIB, Venter E, Avrova AO, et al. 650 An ancestral oomycete locus contains late blight avirulence gene Avr3a encoding a

21 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

651 protein that is recognized in the host cytoplasm. Proceedings of the National Academy of 652 Sciences of the United States of America. 2005;102(21):7766-71. doi: 653 10.1073/pnas.0500113102. 654 33. Huang S, van der Vossen EAG, Kuang H, Vleeshouwers VGAA, Zhang N, Borm 655 TJA, et al. Comparative genomics enabled the isolation of the R3a late blight resistance 656 gene in potato. The Plant journal : for cell and molecular biology. 2005;42. doi: 657 10.1111/j.1365-313X.2005.02365.x. 658 34. McDowell JM, Simon SA. Recent insights into R gene evolution. Mol Plant 659 Pathol. 2006;7(5):437-48. Epub 2006/09/01. doi: 10.1111/j.1364-3703.2006.00342.x. 660 PubMed PMID: 20507459. 661 35. Meyers BC, Dickerman AW, Michelmore RW, Sivaramakrishnan S, Sobral BW, 662 Young ND. Plant disease resistance genes encode members of an ancient and diverse 663 protein family within the nucleotide-binding superfamily. The Plant journal : for cell and 664 molecular biology. 1999;20. doi: 10.1046/j.1365-313X.1999.t01-1-00606.x. 665 36. Hulbert SH, Webb CA, Smith SM, Sun Q. Resistance gene complexes: evolution 666 and utilization. Annu Rev Phytopathol. 2001;39:285-312. Epub 2001/11/10. doi: 667 10.1146/annurev.phyto.39.1.285. PubMed PMID: 11701867. 668 37. Dong S, Yu D, Cui L, Qutob D, Tedman-Jones J, Kale SD, et al. Sequence variants 669 of the Phytophthora sojae RXLR effector Avr3a/5 are differentially recognized by Rps3a 670 and Rps5 in Soybean. Plos One. 2011;6(7):1-8. 671 38. Anh VL, Anh NT, Tagle AG, Vy AG, Inoue Y, Takumi S, et al. Rmg8, a new gene 672 for resistance to Triticum isolates of Pyricularia oryzae in hexaploid wheat. 673 Phytopathology. 2015;105(12):1568-72. 674 39. Varden FA, Saitoh H, YOSHINO K, Franceschetti M, Kamoun S, Terauchi R, et 675 al. Cross-reactivity of a rice NLR immune receptor to distinct effectors from the blast 676 pathogen leads to partial disease resistance. bioRxiv. 2019:530675. doi: 10.1101/530675. 677 40. Guo L, Cesari S, de Guillen K, Chalvon V, Mammri L, Ma M, et al. Specific 678 recognition of two MAX effectors by integrated HMA domains in plant immune receptors 679 involves distinct binding surfaces. Proceedings of the National Academy of Sciences. 680 2018;115(45):11637-42. doi: 10.1073/pnas.1810705115. 681 41. Białas A, Zess EK, De la Concepcion JC, Franceschetti M, Pennington HG, 682 Yoshida K, et al. Lessons in effector and NLR biology of plant-microbe systems. 683 Molecular Plant-Microbe Interactions. 2017;31(1):34-45. doi: 10.1094/MPMI-08-17- 684 0196-FI.

22 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

685 42. Andrivon D. Races of Phytophthora infestans in France, 1991–1993. Potato 686 Research. 1994;37(3):279-86. doi: 10.1007/bf02360520. 687 43. Pilet F, Pellé R, Ellissèche D, Andrivon D. Efficacy of the R2 resistance gene as 688 a component for the durable management of potato late blight in France. Plant Pathology. 689 2005;54(6):723-32. doi: doi:10.1111/j.1365-3059.2005.01288.x. 690

23 bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2020.10.08.331181; this version posted October 8, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.